The effectiveness of soil dispersion procedures is crucial for a successful soil particle size fractionation. In standard methods of analysis of particle size, samples may be exposed to chemical pretreatment to remove organic matter, sesquioxide and other cementing agents before being dispersed by a combination of chemical and physical means (Gee and Bauder, 1986). Chemical dispersion treatments are not considered feasible for isolation of intact organo-mineral fractions.
Most studies of particle size fractionation have relied on ultrasonic dispersion of soil. Several researchers have compared the conventional chemical method with ultrasonic dispersion and concluded that ultrasonic dispersion was as effective as the chemical method (Healy and Oaridge, 1974; Anderson et al., 1981; _Curtin et al., 1987).
2.6.2 Chemical characteristics of particle size fractions
Some of the reported chemical characteristics of particle SlZe fractions are summarized in Table 2.3. It has been shown that the clay fraction often accounts for
50%
of the whole soil carbon (Anderson et al., 198 1 ; Christensen, 1985; Curtin et ai.,1 987; Balesdent et al., 1988; Bonde et al., 1992). To compare the amount of soil
organic matter in a particle size fraction from different soils, the carbon enrichment ratio (mg carbon/g separate to mg carbon/g whole soil) waS calculated for each fraction. Some of the reported values in the literature are summarized in Table 2.3. Among the size separates, clay and silt had substantial amounts of amorphous and
crystalline Fe and AI oxides (Table 2.3). The substantial amount of organic matter
and sizable portion of Fe and AI oxides present in the clay and silt fractions showed
the importance of these fractions in relation to other aspects of soil chemistry, such as the retention and cycling of nutrients, pollutants and pesticides.
2.6.3 Contribution of organo-mineraJ fractions of soils to sorption of pesticides
Contamination of groundwater and surfacewater by soil-applied pesticide residues is largely controlled by the extent of their sorption onto soil particles. While the sorption of a pesticide onto soil particles decreases the leaching of pesticide residues
Table 2.3 Summary of literature on the chemical characteristics of particle size fractions
::!jl£����cif:'Cban;Cteiis:���:(��t:��:'�D� •.••••••• ·.'."
4 Soils - fine claySaskatchewan, coarse clay
Canada fine silt medium silt coarse silt sand 10 soils- New Zealand 1 soil- Turkey 1 soil - USA 7 soils - New Zealand 2 soils - Saskatchewan - Canada 2 soils - Saskatchewan - Canada clay silt sand fine clay coarse clay fine silt medium silt coarse silt fine clay coarse clay medium silt coarse silt sand clay silt sand fine clay coarse clay fine silt coarse silt sand fine clay coarse clay fine silt medium silt coarse silt sand 4 soils - Canada fine clay
coarse clay fine silt medium silt coarse silt
Soil carbon associated with < 2 Ilm, 2-50 Ilm and > 50 Ilm fractions averaged 53.5, 43.4 and 3.3 %, respectively. Significant amounts of pyrophosphate and citrate dithionite - bicarbonate Fe and AI were present in clay and silt fractions. Carbon enrichment ratio
was 1 .6-2.9 and 1 .4-2.4 in fine and coarse
Oay and silt fraction has substantial amount of carbon and CEC.
Citrate dithionate and oxalate extractable Fe and AI decreased with increasing particle size.
The organic carbon content increased with decreasing particle size.
Oay and silt fractions contained the largest proportion of whole soil carbon.
55 to 58 % of organic carbon was in the clay
fractions, with the greatest amount in the coarse clay (2-0.2 urn) fractions. The carbon
enrichment ratio for fine silt, coarse clay and fine clay samples was 3.8, 2.25 and 1 .8
respectively.
Oay, fine and coarse silt contained substantial amount of citrate dithionate bicarbonate extractable Fe and AI.
Significant amount of carbon was present in coarse clay, fine silt and medium silt. Ho\Vever fine silt, fine clay contained substantial amount of amorphous Fe and AI.
Curtin et al. ( 1 987) Healy and Oaridge ( 1 974) Sayin et al. ( 1990) Nkedi-Kizza et al. ( 1983) Tate and Churchman ( 1 978) Anderson et ai. (1981) Huang et a/. ( 1984) Schnitzer and Kodama ( 1 992)
Table 2.3 continued
1 soil - Canada clay Highest organic matter was present in clay Kay and
silt (10.61 %) followed by silt (2.51 %) fraction. EIrich (1967) sand
4 soils - clay Sand size fraction contained less than 10 % Christensen
Denmark silt of the soil carbon, whereas the clay contained (1985)
fine sand 1 49-69 % and silt from 22-38 %. The carbon
coarse sand enrichment factor ranged from 3.3 to 16.4 for
fine sand 2 clay and from 1 .8-9.9 for silt fraction.
6 soils - clay The clay « 2 urn) contained about 50 % of Dalal and Australia silt the soil organic matter, whereas silt and sand Mayer (1986)
sand had about 25 % each .
• Fine clay: < 0.2 .um; coarse clay: 0.2-2 .urn; fine silt: 2-5 .urn; medium silt: 5-20 .urn; coarse silt: 20-50 .urn; sand: > 50 .urn
to groundwater, the erosion of pesticide enriched finer soil particles by runoff leads to surface water contamination.
It has been well documented that eroded sediments are enriched with clay size fractions and organic matter (Leonard, 1990). As discussed in section 2.4.1.1, a large volume of work has been carried out on the sorption of pesticides by whole soils, a limited amount of work has been reported on the effect of particle size on pesticide sorption (Table 2.4). Most of these studies have concentrated on the sorption of pesticides and desorption has not been reported in the organo-mineral fractions.
In New Zealand, sheet eroSIOn is considered to be the dominant erosion type affecting approximately 217400 and 8318400 ha of land in the North and South Islands, respectively. Pesticide runoff includes dissolved, suspended particulate and sediment sorbed pesticide that is transported by water from a treated land surface (Leonard, 1990). Grant (1992) reported that one hectare could produce 6.4-10.3 tonnes of sediment from allophanic and non-allophanic soils of New Zealand at a rainfall event of 65mm. Some of the reported losses of pesticide during natural and simulated rainfall events are summarized in Table 2.5.
Approximately 1-2% of applied pesticide was lost through natural rainfall events, whereas losses of 5-10% were reported during simulated rainfall. Pesticide contamination in water bodies was much less than that measured in sediments.
The major source of pesticide contamination of groundwater and surface water is through the movement of pesticide residues in dissolved organic carbon (DOC) in soil. The importance of DOC in the sorption and movement of pesticide residue is discussed below.
2.7 EFFECT OF DISSOLVED ORGANIC CARBON (DOC) ON SORPTION AND MOVEMENT OF PESTICIDES
As discussed in section 2.5.3.1, soil organic -matter is the principal sorbent of the majority of pesticides. In recent years, DOC has been the subject of considerable interest, because it has been reported to interact with pesticides, and thus affect the
Table 2.4 Soil: sand coarse silt medium silt fine silt coarse clay medium clay fine clay Soil: sand silt clay Soil sand silt coarse clay medium clay fine clay clay Soil: sand coarse silt medium silt fine silt coarse clay fine clay Soil: sand coarse silt fine silt clay
Summary of literature o n particle size contributions to pesticide sorp tion-desorption by soils and sediments
Diuron and Adsorption increased with Nkedi-Kizza
2,4,5-T decreasing particle size in et al. ( 1 983) proportion to OC content.
Freundlich K values vary within a factor of 7, whereas the K"., values vary only within a factor
of 1 .5 . The K"., values for non-
ionic (Diuron) on whole soil and particle size are essentially the same. For ionic (2,4,5-T)
the clay, silt and whole soil are
the same and it is three times lower for the sand fraction.
Lindane Increased adsorption of lindane Kay and EIrich
has been observed for clay (1967)
followed by silt and sand
fraction which corresponds to
an increase in OC content.
DDT, Retention of pesticides was Richardson and
Methoxychlor highest in the natural and Epstein (1971) Endosulfon organic matter removed fine
and very fine clay fractions.
Removal of OM with H202 reduced the retention of
pesticides in the fine clay
fractions.
Atrazine Besides organic matter, Huang et aL
noncrystalline AJ and Fe (1984) provide sites for atrazine
adsorption, particularly in the
< 20 .um fractions.
Tribunil Adsorption increased with Raman (1987) decreasing particle size. The
Freundlich K value varied by a factor of 3 or 4 but K"., values were quite similar among the
Table 2.4 continued
Sediment: Aromatic and The partition coefficient (KJ Karickhoff
sand chlorinated were independent of sediment et al. ( 1979) coarse silt hydrocarbon concentration and directly
medium silt related to OM content. The
fine silt
�
increases with increasingclay sorbent OC content. Highest
K"., of compounds for medium
and fine silt followed by clay
fraction and the lowest in sand
fraction.
Sediment: Polycyclic Positive relationship observed Evans et al. (1990)
clay and silt aromatic between % OC and P AH held
sand hydrocarbons by the sediment.
(PAH)
Sediment: Paraquat Paraquat concentration is 20-50 Karickhoff and
sand fold higher in clay or fine silt Brown (1978)
coarse silt compared to medium and
medium silt coarse silt and as much as 1000
fine silt fold higher than sand.
coarse clay Isotherm coefficient
medium clay determined on individual
fine clay particle size fractions can be used to compute the paraquat·
distribution as a function of
particle size in the whole sediment .
• Fine clay: < 0.08 .urn; medium clay: 0.08-O.2 .um; coarse clay: 0.2-2 .um; clay: < 2 .um; fine silt:
Table 2.5 Atrazine Alachlor Picloram Glyphosate 2,4-D Picloram 2,4-D Cyanazine Atrazine Simazine Atrazine Propachlor A1achlor Atrazine Carbofuron Atrazine Atrazine Alachlor Chlorpyrifos A1achlor Carbofuron Terbufos
Summary of literature on the runoff losses of pesticides from natural and simulated rainfall
.. ··• Maryland Maryland Arizona Ohio Oregon Oregon Saskatchewan, Canada Pennsylvania Maryland Iowa Iowa Iowa California Stoneville, USA Wisconsin Illinois
(:��pjcci�er.:. ..
: :::;"::.
.. . .
of apphcahom
<
.. Natural rainfall Com 0.1 Com 0.16 Pinyon-Juniper 1 .1 No-till com < 1 Rangeland 0.014 Rangeland 035Wheat stubble, 4.1 (6-year average)
fallow 03 Fallow 0.73-5.7 conventional 0.01-0.75 (no-till) Com 1.6 (Conventional) ( conventional 1.1 (no-till) & no-till) Simulated rainfall Corn 0.76-6.1 0.97-5.7 1.0-8.6 Fallow and 0.7-6.3 corn residue 0.8-11.4 Corn 4-8 (emulsion liquid) 9-12
(granule and wettable
powder) Grass 5.8 4.0 0.2 Corn 1-2 1-1 1 1-6
." >
·
••.. Wu (1980) Johnsen (1980) Edwards et al. (1980) Norris et al. (1982) Nicholaichuk and Grover (1983) Hall et al. (1984) Glenn and Angle (1987) Baker et al. (1982) Haith (1986) Wauchope (1987) Sauer and Daniel (1987) Felsot et al. (1990)fate of these pesticides in the soil or aquatic system (Ballard, 1971; Carter and Suffet, 1982; Landrum et ai., 1984; Madhun et ai., 1986b; Lee and Farmer, 1989). Several investigators have reported that the apparent solubilities of pesticide solutes increase in the presence of DOC (Wershaw et ai., 1 969; Chiou et ai., 1986), and this can sometimes increase the movement of pesticides in soils (Kan and Tomson, 1990; Logan et ai., 1992; Liu and Amy, 1993).
In order to understand the mechanisms involved in the interaction between DOC and pesticides, it is necessary to know the sources and characteristics of the DOC.
2.7.1 Sources or DOC
The sources of DOC are grouped into endogenous and exogenous DOC and are briefly discussed below.
2.7.1.1 Endogenous DOC
This consists of naturally-occurring dissolved material namely humic acid, fulvic acid and water soluble soil organic matter. These are macromolecules with complex three-dimensional molecular structures comprised of hydrophobic and hydrophillic sites (Wershaw, 1986). It has been shown that the composition of water soluble soil organic matter is similar to that of humic and fulvic acid (Linehan, 1977).
2.7.1.2 Exogenous DOC
A number of organic manures, which act as a source of organic matter are applied to soil through agricultural practices. These include: sewage sludge, pig manure, mushroom compost, poultry, cattle and sheep manure. It has been shown that application of manure greatly increases the amount of soluble carbon in the soil solution (Meek et ai., 1974). Studies examining addition of exogenous DOC on the fate of pesticides are scanty (Barriuso et ai., 1992b).
2.7.2 Measurement of DOC
Various methods have been used to measure the DOC in soils. These include: (i) a dry combustion method, which involves combustion of the sample at a high temperature
(>
300°C) with subsequent measurement of the evolved CO2 (Bremner and Tabatabai, 1971); (ii) A wet oxidation method in which the amount of oxidising agent used in the process is measured (Tate et al., 1988); and (iii) The absorbance of light at different wavelengths (250-300 nm)
for measuring DOC concentration inwater samples (e.g., Dobbs et al., 1972; De Haun et al., 1982; Moore, 1985; Timperley, 1985).
2.7.3 Effect of DOC on sorption of pesticides
Several studies have suggested that sorption of pesticides by soil colloids or sediments is affected by the presence of DOC in the soil solution (Carter and Suffet, 1982; Lee and Farmer, 1989). Hassett and Anderson (1982) observed that the addition of DOC derived from natural water and sewage decreases the sorption of hydrophobic organic compounds. O'Connor and Connolly (1980) proposed that the observed decrease in the partition coefficient of hydrophobic compounds with increasing sediment concentration can be explained by the release of DOC from the sediment to the
aqueous phase. Similarly Lee et al., (1990) observed a reduction in sorption of napropamide on clay when peat humic acid was present in the slurry.
Caron et ale (1985) found that the addition of DOC to the aqueous phase reduces the
sorption of DDT by sediments. Recently, it has been shown that DOC from exogenous and endogenous sources influenced the sorption of pesticides in soil
(Barriuso et al., 1992b). They reported that preincubation of soil with DOC generally
increased sorption, whereas pesticide sorption decreased when DOC was preincubated with pesticide.
Several reasons have been attributed to the decrease in sorption of pesticides in the presence of DOC. These include: (i) interaction between DOC and pesticide in solution (Carter and Suffet, 1982; Madhun et al., 1986b; Lee and Fanner, 1 989), and (ii) competition on sorption sites between DOC and the pesticide (Lee et al., 1990).
Infrared (IR) spectroscopy has been shown to be a good technique to reveal the binding mechanisms of pesticides with DOC (Burns et al., 1973; Khan, 1 974; Senesi and Testini, 1980; Madhun, 1984). The binding mechanisms include ion exchange process in diquat, paraquat, chlordimeform and triazines (Madhun, 1984; Burns et al.,
1973; Khan, 1973, 1974; Maqueda et al., 1983; Senesi and Testini, 1980), hydrogen bonding in triazines and substituted ureas (Senesi and Testini, 1980, 1983; Madhun, 1984), and charge transfer complexes in triazines, substituted ureas, and bipyridilium herbicides (Khan, 1973, 1974; Mueller-Wegener, 1977; Senesi and Testini, 1 980, 1982, 1983; Madhun, 1 984). Recently, Hermosin and Cornejo (1993) studied the binding mechanism of 2,4-D by organo-clays using IR spectroscopy and reported a hydrogen
bonding between carbonyl group of 2,4-D and ammonium group of the interlayer organic cations.
2.7.4 Effect of DOC on movement of pesticides
In recent years, the effect of DOC on the transport of pesticides has attracted much attention because of (i) the interaction between the contaminant and DOC and (ii) mobility of DOC in the aquifer. The literature pertaining to the effect of DOC on the transport of pesticides is briefly discussed below.
Abdul et al. (1990) reported that humic acid is more effective than water in removing the more hydrophobic nonpolar organic contaminants· from aquifer material. Similarly Logan et al. (1992) showed the enhanced movement of chlordane in the presence of humic acid. Kan and Tomson (1990) and Liu and Amy ( 1993) also demonstrated the facilitated transport of organic contaminants in the presence of DOC.
Recently, it has been shown that in soil column studies aqueous phase DOC enhances the transport of polynuclear aromatic hydrocarbons, whereas solid phase DOC retards transport (Liu and Amy, 1993). Enfield et al. (1989) demonstrated the facilitated or enhanced transport of the hydrophobic compounds pyrene and hexachlorobenzene in the presence of DOC ..
concen tration in soil solution thereby decreasing pesticide sorption (Lee et al., 1 990). Addition of liming matelials, ammonia or other basic fertilizers is veIY effective in increasing soil pH. The increase in soil pH may enhance pesticide concentrations in the solution phase, thereby contributing to greater mobility in the soil. Smith and Willis ( 1985) demonstrated enhanced movement of several pesticides in soil columns with the application of anhydrous ammonia.
2.8 MOVEMENT OF PESTICIDES IN SOIL
The movement of pesticides through soils following landfill disposal or agricultural application is of great concern whenever there is potential threat to groundwater contamination. The miscible displacement technique is commonly employed in the laboratOlY using columns of repacked soil to examine the physical and chemical processes involved in the movement of solutes.
Two types of experiments are commonly conducted to examme the mechanisms involved in the leaching of pesticides. The first and most common type involves a step-function change in influent concentration (CJ from 0 to Co for the chemical species of interest during the transport of solute through a soil column. The leachate concen tration (Ce) is monitored following this change, and a breakthrough cu rve (ETC) is obtained relating CjCo to the number of liquid-filled pore volumes (p) of solute. Often Cj is then changed back to 0 and the falling limb (or desorption) of the ETC measured.
In the second type of experiment, a pulse of solute is applied to the soil surface or mixed with the surface soil. Water is then usually percolated through the column for a period and the d istribution of applied solu te within the soil measured by destructive sampling.
2.8.1 Modelling solute J eaching
Models are a simplified description of reality. The development of simulation models for forecasting pesticide behaviour is an attractive way of evaluating solutions to some agricultu ral and environmental problems. Models are being increasingly used as
management tools to predict the fate of pesticides in soil (Wagenet and Rao, 1990).
2.8.1.1 Uses of simulation models
Development of simulation models, which predict the leaching of pesticides in soils have the following benefits: (i) help to assess the time required for the soil-applied pesticide to be dissipated to some acceptable, regulated level before entering groundwater. (ii) provide information on the likely environmental impact of a new pesticide before its actual use. (iii) assist farmers and growers in designing effective crop, soil and chemical management strategies. (iv) achieve a minimum environmental impact and maximum crop-yield response from the minimum amount of pesticide application.
2.8.1.2 Type of simulation models
Many models for the transport of non-reactive and reactive (pesticide) solutes have been developed over the years, some simple and some complicated. The simplicity or complexity of a model often depends on whether it is constructed from a research, management or screening perspective. Several reviews have been written recently on solute transport models (Addiscott and Wagenet, 1985; Brusseau and Rao, 1990; Jury and Fliihler, 1992; Selim, 1992) and also the various pesticide fate models have been discuSsed in detail (Wagenet and Rao, 1990). The importance of sorption in the movement of pesticides can be examined by using the Convection-Dispersion equation (CDE). This model has been used extensively to simulate the movement of pesticides (e.g., Davidson et al., 1968). The CDE model involves both equilibrium and non-equilibrium sorption processes. Each model is briefly discussed below.
2.8.2 Convection-Dispersion equation
The convection-dispersion equation (CD E) is the classic deterministic and mechanistic approach, based on miscible displacement theory (Nielsen and Biggar, 1962), to modelling the transport of solute through the soil by water. This model assumes that physical convection (Le., mass flow) and molecular diffusion combine to displace a solute in a porus media.
2.8.2.1 Non-sorbing solutes
For steady-state water flow conditions, the non-reactive solute movement can be
described by the CDE as (Nielsen and Biggar, 1962; Kirkham and Powers, 1972; Wagenet, 1983; Cameron and Scotter, 1986)
(2 . 10)
Where t is time (hr), D is the dispersion coefficient (mm2 hr-1), x is distance in the
direction of flow (mm) and v is average pore water velocity (mm hr-t).
The solution of Equation 2.10 for the case of a continuous solute input is,
CICo
=1/2
[eifel . 1 -p]
+ evUD eifel1 +p
D
(2 .11)
2 (Dp/vL)1/2 2 (Dp/vL) 112
Where p is liquid-filled pore volume, L is column length (mm). The relation between